DE1290995B - Scattered beam diversity transmission system - Google Patents

Description

to create system that has only one transmitting or receiving device and one associated with each Antenna gets by and that allows a satisfactory compensation of the fading. At the same time the 5 previously required in the known systems expenditure on transmission power can be reduced.

with ultrashort waves or microwaves straight io transmission channels after a certain time and propagate, over the horizon to over-encode carrier frequency scheme in such a way that the carry. If such waves on atmospheric messages are modulated by means of basic signal pulses irregularities of the dielectric constant, which collide in a number of equal parts, a scatter results. In this way pulses are subdivided, each of which is transmitted on a very small proportion of the energy of a separate frequency. The number of the falling wave according to the laws of chance zer- partial impulses per basic signal impulse and the number of scatters, whereby then the individual frequencies scattering in the forward direction are arbitrary.
Part can be transmitted over the horizon, the invention creates the advantage that a special

by having a suitable antenna on the part which is sufficient to ensure fading suppression Troposphere aligns, in which such a scattering 20 and that also the necessary for transmission occurs. Bandwidth can be kept much smaller

To make practical use of such a system than with previously known diversity systems. Thinking, generally have to be very large on the transmission side, but the individual frequencies remain so far Energies are applied to the very high apart that they are in varying degrees To take into account wastage. It is also necessary to be subject to the extent of fading. Also earthly, Directional antennas with high gain are reduced to required transmission power. Simultaneously To use the energy density in the desired way, the technical effort is hereby considerably forward-looking to enlarge. discontinued, and is also important in terms of

A property of this type of transmission is the portability of the devices, which at the same time is achieved according to the law of chance extinction 30 reduction of their own weight,
and amplification of the received signals as radio. Advantages are for coding and decoding

tion of time. The frequency with which these changes occur in the signals of all transmission channels at the transmitter and genes occur is in the order of 10 Hz. provided on the receiver program selector. For To compensate for the reduction in the size of the signal, each frequency is expediently a separate one strength during the fading times, either a large oscillator and a combinatorial circuit are arranged in Energy on the transmission side is required, or the signals must be transmitted to the antenna before they are transmitted sen receiving and / or transmitting side Mehrfachverfah- according to the time and carrier frequency scheme ren (diversity) are applied. These can be combined multiple times. The coding of the signals in methods that use a number of signals, this scheme is then advantageously carried out by means of in the whose fading characteristics are independent of each other 4 ° channels of the transmitter lying modulators, which are, allow a reduction in the transmit periodic pulses generated based on information power. of the associated channel, and by means of a

The post-generator belonging to the state of the art, which has a light system for each modulator pulse, which generates a plurality of partial pulses with tropospheric scattering, which process for one, used among other things continuous 45 correct information channel each with a certain signal and tried with the help of multiple transmitters show temporal distribution,
and / or receivers with a plurality of from For the execution of the transmission is according to

antennas located in front of one another are statistically independent of one another in a further development according to the invention To achieve signals. It has already been beaten that every channel of the receiver FreandcTc Communication systems, for example systems 5 ° contains frequency dialing devices which receive the Multiple angles or multiple polarization catch frequencies in groups of partial pulses

separate, and that another distribution device it is provided that each group of partial pulses is converted into individual pulses for the individual information 55 separates channels.

Another advantageous property of a transmission system according to the invention is that despite the relatively small distance between the frequencies used for transmission, there is interference

where two pulses of the same frequency rarely occur due to crosstalk. To such a are transmitted one after the other in order to identify the correct disturbance even more strongly if necessary to distinguish thereby from interfering noises, switch, it is proposed that for each of the that the two transmitted impulses in temporal channels of the receiver one with maximum true relation be brought to each other. Such an apparent working detector is provided, Transmission systems can only mix 65 of the between pulses encoded for that channel by switching off messages and background noise, and crosstalk impulses that differentiate that however, do not compensate for fading. have a smaller amplitude than that for this

The object of the invention is to provide a transmission-specific channel coded pulses.

been spiral. However, each of these systems requires at least a separate one for each diversity path Receiver, in other cases several senders and receivers.

They are transmission systems with only one Transmitter and a transmitting antenna and accordingly with only one receiving antenna and one Known receivers for scatterbeam diversity, with

A synchronization between transmitter and receiver is advantageously carried out in such a way that the receiver a timing circuit, a phase detector circuit and a synchronizing circuit, the latter being the The timing circuit is in synchronism with the timing circuit on the transmitter side, and the phase detector circuit and the synchronization circuit respond to the signals of the data channel when synchronizing, the one next to is an additional transmission channel to the channels provided for normal program transmission.

The transmission of voice information takes place with a transmission system according to the invention ideally so that the signals containing the speech information in each channel are divided into step-like partial signals To be sat down.

The invention is explained below with reference to the drawing, for example. In the drawings demonstrate:

F i «. ia and 1b placed side by side, the transmission part of the transmission system, whereby the program connections 1, 2, 19 and 20 and also the other connection lines A to F of the two sheets of drawings are to be thought of as connected to one another,

F i g. 2 shows a representation of typical wave trains of the 19 language programs which are transmitted during a 128- [tsec-Samineleiiode by the slufen converters of the PrograiTiniwäiiierinatrix in a system that is fully utilized,

3a, 3b and 3c three associated representations, FIG. 3a shows the input waveform, its slope conversion and the time-dependent sawtooth voltage represent ι,

F i g. 3 b converting the message into a. inipiiioiagenmodulated message during one of the Scanning periods of FIG. 3,

F i \ :. 3 c the output coding of two of the sprariiprogramrne to explain the individual codes, which are characteristic for the outputs of the individual gate circle oscillators,

Fic. 4 a considerably simplified-.es Pnaz ^ sc ^ a ¹ · - picture of the üii transmitter used I.-latrix. which ctie message receives from the stage conversion irn ^r lt " Sprachprograinme and distributes this message to a multiplex system with five working frequencies,

Fig. 5 is an illustration of the waveforms at Output of the program matrix, which defines the coding pulse distribution and that in FIG. 2 indicated meetings the pulse positions can be recognized, and

6a and 6b, placed side by side, show a block diagram of the receiver with which the message is received from the transmitter according to the invention and distributed over 19 transmitters and data processing groups will.

The in F i g. 1 a and 1 b shown transmitter arrangement works with 19 voice channels, of which as Examples only the voice channels 1, 2 and 19 are given. The 20th channel, hereinafter referred to as the data channel, is generally used to transfer data, but especially to synchronize the Receiver with the transmitter. If there is any data to be transmitted, it will run through this Channel; if none are to be transmitted, this channel only transmits a stream of idle data. The necessary Frequency and phase information is then obtained from the stream of that channel for the required Synchronization process pulled out. In this respect, the 20th channel has two functions, one of which one, synchronization, must always be carried out, while the other is also optional can be.

To operate the 19 language programs, a quantized pulses of graduated size are used working layer modulation scheme used. In order to obtain a larger one within the available bandwidths Number of channels, only 16 pulse levels are used per channel. It is known that for high quality voice transmission

ίο Up to now around 64 pulse levels were required. However, using these multiple levels in a tropospheric dispersion system would significantly degrade performance because of the greater bandwidth required for each speech program. In the system according to the invention, therefore, only 16 pulse level ^are: ·.-Π <ends, resulting in the higher power due date of the transmission channel results. In other words, the savings in bandwidth from using this smaller number of

ao Impiilsgradungen for voice transmission leads to a greater number of available voice channels. To get to the fourth part of the previously needed To enable a qualitatively satisfactory speech transmission for impulse levels becomes a special one Language processing method, which will now be discussed in detail.

The language programs 1, 2 and 19 of FIG. 1 will now be discussed as an example. The first step in the processing of the signals received by the microphones 12 language is '' juenzkreisen that as a preamplifier 13, pre-emphasis circuit 14, fast automatic gain control 15 and transmitter band filter IC are indicated in the results shown in each channel as a rectangle tonf ^r. The indices used in Fig. 1 on the aforementioned and other rectangles serve to identify the corresponding language program. For the sake of completeness, it should also be mentioned that in the case of the f -'sierr? between language programs 2 and 19

·, = 16 more language channels are available.

The audio-frequency skrsai · .. 'from the devices 12 are preamplified in conventional preamplifier circuits 13 and then fed to the preamplifiers 14, in which the signals receive an increasing frequency response of 6 decibels per octave in a known manner.

After the signal has been pre-emphasized, it is fed into the fast-responding automatic gain control circuits which are shown in FIG. 1 are denoted by the reference symbols IS ₁ , IS ₂ and IS _{1 ,,.} The task of these groups is to generate the same average audio frequency level within a useful dynamic range in each individual voice channel, regardless of the volume of the voice source. In this way, all subsequent circuits will receive the correct signal level for proper operation, thus reducing the possibility of overmodulation.

After the automatic gain control circuits 15, the signals come into those already mentioned Band filter to limit the audio frequency signals to the range of approximately 300 to 3000 Hz.

The signals reach the instantaneous value compressor circuits from the audio frequency circles shown as rectangles 13 to 16 17. This processing results in a uniform distribution with a good approximation of the speech amplitudes.

The language processed by the instantaneous value compressor circuits 17 is then shown in FIG. 2 periodically scanned and according to FIG. 3 a converted to quantized signals. These operations take place in the audio frequency sampling and storage circuits 18, the pulse phase modulators 19 and the converters 20 (Fig. la), the following to be discribed.

In this particular embodiment, uniform signal sampling is used. The sampling frequency is approximately 7.8 kHz, circles 18 to 20 of a particular below clock pulse generator 24 described in more detail are supplied corresponding clock pulses.

At the time of sampling, the audio frequency amplitude in the audio frequency sampling and storage circuits

18 saved to a very narrow, about 1 jisec to generate a continuous pulse, the timing of which is a measure of the sampled amplitude (FIG. 3 b). This distribution of the pulses in terms of length takes place in the pulse position modulators 19.

The narrow pulse is generated at the point in time when one increases linearly with time Sawtooth voltage, which begins at the time of sampling, the amplitude of the stored audio frequency Sampling reached. This short pulse therefore assumes a temporal position, that of the amplitude is proportional to the stored sample.

It is now necessary to insert this time value in one of 16 time fields which are between the Audio frequency sample time of the currently stored sample and the time of the next sample lie. It can be seen that this sampling period is exactly 128 μβεΰ. The value of 128 nsec is the reciprocal of the sampling frequency of 7.8 (exactly 7.8125) kHz. Thus, each of the 16 line fields has a length of exactly 8 (isec. With the selected Sampling frequency of 7.8 kHz is a value that has a good relationship between the Ensures pulse width (bandwidth) and the quality of the sampled speech.

The pulse position is implemented in the

19 voice channels with the help of modulators 2O ₁ to 2O ₁₉ . It should be mentioned at this point that the output of the converter which is fed to the program selector matrix 23 of FIG. 1b is a sequence of 8 nsec pulses, one pulse occurring in each of the successive 128 jisec sampling periods . The special position of these cS-iiscc pulses within one of the 16 specific 8- [isec time fields in each 128 nsec sampling period embodies the amplitude of the sampled audio-frequency signal. The pulse train thus forms a converted pulse position modulation pulse stream in each of the speech channels used.

The signal pulse modulated by the pulse position modulated in each period (hereinafter PPM signal pulse = named pulse position modulation) the field occupied changes according to the law of chance free of one voice channel to the other. This change is desirable in order to make the available energy more even over time to distribute and to reduce the maximum instantaneous power required by the transmitter.

F i g. Figure 2 shows a typical pulse scheme of 19 speech channels. as it is supplied by the converters 2O ₁ to 2O _1n to the program selector matrix 23 in a fully utilized system. The time between ocn base timestamps / "to / _; , corresponds to the 128 nsec sampling period. There is an 8 ^ sec part-time field between the time marks t _c and t _d. A pulse occurring in any given speech program appears in any of the 16 separate 8 ^ sec part time fields of the 128 | xsec sample period t _u through t _b . As the language lines 1 to 19 of FIG. 2 show, these partial pulses occur completely randomly in the individual voice channels. For example, fields 6 and 11 can remain free, while fields 2, 7 and 9 contain two pulses and field 10 three pulses.

For the speech program 13, for example, the PPM pulse occurs in field 12. This point resulted from the special momentary speech amplitude of this program at the sampling time t _a . It is from the fig. 2 that each language program fed to the program selector matrix 23 contains an impulse for only one sixteenth of the time = 8 seconds. Due to the statistical

ao distribution of the speech amplitudes, the impulses from line to line are arbitrarily in an approximately uniform Way distributed.

The various pulse current outputs of the language programs supply the program selector matrix 23 (Fig. 1 b), which also contains a sequence of time pulses from the time pulse generator 24 (Fig. 1 a) receives. The latter, delivers direct current pulses with a width of 1.33 (isec. A stable time circuit 33 with a frequency of 75OkHz supplies the time pulse generator a base time signal. A time sequence of six of these 1.33 μβεο long pulses falls in each case combined with one of the 8 nsec impulses that are applied to the converted in the manner described above PPM pulses are due. The time pulse generator also generates a synchronizing pulse, which always marks the beginning of such a group of sets six 1.33 fisec pulses. This last mentioned Information is fed to the converters 20.

The time pulse generator 24 preferably works with ring counters. In addition to the normal way of working, which consists of generating 1.33 | isec time division pulses t _t to / "with the help of the 750 kHz clock 31, the ring counters have the task of generating series of 8 ^ sec pulses, which occur one after the other exactly in the 16 time fields that define the implemented pulse positions.

Lines going out from the various points of the ring counters are shown at the upper edge of the time pulse generalor 24. These lines are arranged so that an 8 nsec pulse is available _{on line F 1 for time field 1 of the sample period.} Similarly, the pulses for the remaining time fields are available on lines that have the same indices as the corresponding fields.

In the embodiment according to FIG. 1 a, the pulses F ₁₅ and F _{10 of} the fields 15 and 16 were used in a manner to be described below for the audio-frequency sampling and storage and for the pulse position modulation, while the other

So their lines have no special tasks to perform in this embodiment. It should however, it should be mentioned that certain modified embodiments of the invention also extend these further Use pulses, for example, to graduate the sampling times of the various language programs

can to load the system even more evenly.

Each of the converters 20 generates on his own

Output an 8 nsec pulse. which one with one of the

16 8 nsec wide time fields of each sampling period coincide. The leading edge of a 1.33 ^ sec pulse from the time pulse generator 24, which occurs in the time column t _L , serves to mark the beginning of such an 8 μβεο wide field to ensure that the above requirements are met.

In connection with the F i g. 3 c this working method is discussed in more detail below.

The 1.33 ^ sec pulses therefore subdivide each of the 8 usec basic pulses of each speech program into six partial pulses, each of which has a duration of 1.33 μβεο. Since all 8 ^ sec basic pulses from the time pulse generator 24 are fed to the lines. <4 to F of the program selector matrix, six outputs t _i to t _ß , a pulse of 1.33 μβεο duration for the definition of six time columns is generated at each of the six outputs.

The program selector matrix 23 mixes each PPM pulse when it arrives from a voice channel with the six time-division pulses t _t to t ₆ from the time pulse generator 24, so that the leading edge of the 8 μβεο long PPM pulse of a converter 20 is timed with the leading edge of the first l, 33 ^ sec time division pulse, which is referred to below as t _x , and the trailing edge of the PPM pulse coincide in time with the trailing edge of the time division pulse t ₆ . This coincidence in time results from the fact that the leading edge of the partial pulse Z _{1 is} decisive for the initiation of the S ^ sec PPM pulses of each converter. As FIG. 1 a shows, a line leads from the terminal t _x to each of the converters.

From Fig. Ib it can be seen that at the output of the program selector matrix 23 five lines are provided which feed five gate-circuit crystal oscillators 25. The oscillators generate five different frequencies Z ₁ to Z ₅ , the frequency Z ₁ for example 140 MHz, the frequency Z ₂ 141.5 MHz, the frequency Z ₃ 143 MHz, the frequency Z ₄ 144.5 MHz and the frequency Z ₅ 146 MHz.

The coincidence between one of the 8 ^ sec pulses from one of the speech converters 20 with the six time-division pulses from the time pulse generator 24 has the consequence, due to the specially selected design, that a certain sequence of these l , 33 ^ sec pulses occurs. As in FIG. 1 b shown, could e.g. B. with a certain combination in the first time column a / j pulse, in the fifth time column a / ₂ pulse, in the third time column a / ₃ pulse, in the sixth time column a / ₄ pulse and in the fourth time column a Z ₅ pulse occurs, whereas no signal pulse occurs in the second time column. This is the coding of language program 1. The sequence mentioned above is intended for this program only. The receiver to be described below recognizes this time frequency sequence as that of the speech program 1 in order to feed this to the associated output, at which all other frequency time sequences are rejected. The combination of the six time columns with the five frequencies enables 20 different programs if it is assumed that no more than one time column and one frequency are to be used in any two language programs at the same time and if a program is still required which does not have any of the time frequency combinations with any one another program shares. The 20-channel program is used for data transmission.

In the receiver (Fig. 6 a and 6 b) are for each of the A decoder and an output are provided for 20 programs. The coding is explained in more detail below discussed.

The five outputs of the gate circuit oscillators 25 (FIG. 1 b) are combined in an adder 26.

ίο As shown in the fig. To recognize 3 c, that insists The reason for the impulse flow of all language programs at the output of the adder is a signal from a Sequence of five high-frequency oscillations lasting 1.33 μβεΰ with different frequencies. This group of five occurs in a single 8 ^ sec time field on. For each language program is in each of the successive 128 ^ sec sample periods only one field available. When operating all programs, each of the successive 8 ^ sec fields according to the law of chance distributes a signal composed of five frequency oscillations whose time-frequency coding identifies the special program which is used for this Signal is responsible. When the system is fully utilized and all programs are running, it is there is a very high probability that signal oscillations will occur in each of the time fields and that occasionally Time fields that contain impulses from two or even three programs. It is here, however emphasizes that at most only one field in every 128 ^ sec sample period belongs to a program.

The operating frequency of the gate circuit oscillators 25 can, for example, as already mentioned above, be in the 140 to 150 megahertz range. The tropospheric scattered radiation system can be operated in the ultra-short wave or also in the microwave range. For this reason it is necessary to bring the output signal of the adder 26 to the desired output frequency in a conversion circuit. The converter 27 with the conversion oscillator 28 is suitable for this, which can be an oscillator with great stability, which after conversion with the output of the adder generates a frequency component of the desired carrier frequency at the output of the converter 27. A driver stage 29 increases the level of the mixer output sufficiently to operate the power amplifier 30 and achieve the required output level. The required output depends on a number of parameters, for example on the transmission distance, the frequency, the geographical conditions and the losses on the transmission path. For example, an output of 1 kilowatt can be sufficient for operation over medium distances of several hundred kilometers. A suitable frequency generated by the combination of adder and oscillator can be in the range of 5 gigahertz.
The output of the power amplifier 30 feeds the antenna 32 via a duplex arrangement 31. The duplex arrangement allows the same antenna 32 to also be used for receiving the duplex system. The separation between transmitter and receiver is made possible with a protective frequency band. For example, the transmitter can operate on a frequency of 4.95 GHz and the receiver on a frequency of 4.8 GHz. The duplex arrangement has a sufficient one on the transmission frequency

909 512/1344

Attenuation around the receiver input circuits against damage when the Protect transmitter.

The antenna 32 can be a parabolic reflector antenna whose beam width is limited to a minimum value in order to obtain a sufficiently narrow mean bandwidth, which is required to ensure the decorrelation between adjacent multiple frequencies. The beam width is determined from the required special transmission distance or the required Scattering angle.

For the coding in this tropospheric scattered radiation transmission system, it is desirable to use in to use pulse position modulation as effectively as possible. Since the transfer of a special language program only requires a pulse for one sixteenth of the entire sampling period, there is enough time left for the transmission of further language programs. Thus, impulses can be such other language programs are distributed in such a way that they are temporally interleaved. So that the Receiver separate the different impulses and on the corresponding separate receiver voice outputs can distribute, a special coding must be used. In the present case the system includes five frequencies, the main task of which is a high-grade multiple transmission to achieve. Preferably, according to the present invention, the same five frequencies are added used for coding in order to separate the voice channels from one another in the manner described above separate. The order in which the five Frequencies transmitted vary.

This coding is easiest to understand from a consideration of the following table 1. As already mentioned before, each of the 8 ^ sec impulses is divided into six partial impulses, each 1.33 μ8εο duration. The six partial pulses t _x to t ₆ are used in table 1 as headings for the columns. In this way, the combination of frequencies can be read off directly for each of the 20 language programs set up one below the other. Among the combinations of the six partial pulses and the five frequencies there are five programs in which no pulse-time combinations are repeated. These are the five program numbers 4, 8, 12, 16 and 20. If the coding were restricted to these five programs, no cross-talk could occur, ie the pulses of one program could never appear in another program in the receiver. However, it is desirable to make the best possible use of the system. In this particular embodiment, therefore, it appeared more expedient to use additional coding points which allow a maximum of one time-frequency pulse between any two programs at the same time. This change enables a total of 19 such programs plus one program whose time frequency coding stands on its own. In Table 1, programs 1 to 19 are combinations in which a maximum of one common pulse can occur, while program 20, which does not have any overlaps, is intended for the transmission of data and synchronization information, which will be explained later .

Table 1

H H H H H U 1 H H U U U 2 /1 H / 2 U U 3 /1 U H / 3 H 4th / 5 /1 4th / a U 5 / a / 4 H fs 6th / 5 / 3 H H fo / 3 7th / 4 U H U / 3 8th U U / 2 9 U U / 4 H / 3 / 2 10 / 4 / 5 U ft / 4 11 / 2 U ft / 5 / 2 12th / 3 U H ft / a 13th U H / 4 ft / 4 14th U U ft 15th H H / 3 / 2 ft ft 16 U U / ₄ H ft 17th H U / 2 ft 18th U / 4 U U ft 19th U H U U U 20th /1 / 2 / a U

For a slightly different explanation of the coding method, reference is made to Table 2, in which the individual columns correspond to _{the output frequencies Z 1} to f _B. The time column in which each frequency appears for a given program is identified in the special frequency column of this program by the time column number. The F i g. 3 c shows at L how the frequencies Z ₁ to f _s are generated in the special time sequence for programs 1 and 2 according to Table 2.

Table 2

/ ι H H H H 1 k H H H H 2 H H H H H 3 H H si " H H 4th H H si " ti ' 5 si = H "1 si " H 6th k si " si " ti * H 7th H H si = H H 8th H H ir, si ' H 9 H H H H H 10 H H H k 11 H H si " H ts 12th ti ' H H si = H 13th k k H H si " 14th ti ' H te, H 15th H H ti ' H 16 H * 2 H H H 17th H H si " k H 18th H H H si " H 19th H H H si " 20th H H H si " si =

The various previously described codings are made with the aid of the program selector matrix 23 generated, at the five outputs of which the special frequency release pulses appear. The impulses on these five outputs are fed to the input of the corresponding gate circuit oscillators 25. In order to describe the mode of operation of a gate circle oscillator 25, we shall now turn to details of such Oscillator.

This oscillator works, for example, in the 140 MHz range; it is designed in the usual way, uses a quartz crystal, which is kept at a constant temperature with a heating device for reasons of frequency stability, and contains a transistor as an active oscillating element. This oscillator generates a small output voltage that is picked up via a gate circuit. Normally the oscillator is short-circuited with the gate circuit to ground in such a way that no high-frequency output appears at the input of the adder 26. With the help of a release pulse from the program selector matrix, the gate circuit is opened in order to allow the high-frequency output of the oscillator to appear at the input of the adder. A sequence of five release pulses, which corresponds to a special program of the five corresponding program selector output lines, switches on the five gate circuit oscillators 25 with the frequencies f ₁ to / ₅ in the correct chronological order. As an example of a special code, language program 1 is selected from FIG. 1b in order to explain the aforementioned circuit sequence in more detail.

With reference to the FIG. 1 b specified at the output of the program selector matrix Pulse code will now be explained in more detail with reference to FIG. 5 of this scheme.

The language program 1 of Fig. Ib can in Field 2 of FIG. 5 can be found again and is identified there with the reference numeral 1. at the representation of FIG. 5, however, is a general representation of the system if this is used with all 19 programs. For this reason it can indeed be given in a given Field more than one group of pulses appear. Such further impulses belong to others Language programs. For example, field 2 of FIG. 5 also an impulse group from the language program 9.

As will be described later, the RF pulses from the gate circuit oscillators 25 are in the Adder 26 is added. This adding process is shown in FIG. 5 indicated as amplifier output power level. From this illustration it can be seen that the RF energy level at the amplifier output changes in individual steps during this typical sampling period according to the law of chance.

Further language programs are shown in FIG. 5 labeled with their corresponding program numbers. The in F i g. Speech program pulses illustrated in FIG. 5 are those previously used in FIG. So puts the FIG. 5 shows the original speech program pulses of FIG. 2 after going through the program selector matrix 23. FIG. Fig. 5 also shows the law of chance overlapping of the time pulses of the different channels and their distribution over the time columns.

With regard to this overlap of the pulses, a case is shown in field 2 in which two of the gate circuit oscillators are switched on at the same time. For example, in the first time column of field 2, program 9 switches on an oscillator f ₅ and program 1 _{switches on an oscillator Z 1} . For this reason, the level shown below for the power output is doubled. This ultimately results in the transmission of a peak pulse energy that is twice as large as compared to the transmission of a single pulse. A

A similar example can be seen in time field 10, in which three oscillators are switched on at the same time. The second time column of field 10 shows that program 7 is oscillator / ₂ , program 17 is oscillator / ₃ and program 4 is

Oscillator / ₅ switches on. As a result, the power output is three times that of a single pulse.

The F i g. Fig. 5 also explains another phenomenon, namely that of pulse division. For example,

ao th in column 2 of field 7 the program 18 and also the program 15 the oscillator / ₄ . Understandably, this coincidence only results in a single RF transmission on the oscillator / ₄ . As a result, the power output corresponds to the level of a simple pulse. This situation is known as energy sharing. A statistical analysis shows that such an energy division results in a small saving in average energy requirements.

An even better understanding of the above statements results from a consideration of FIG. 3 a to 3 c, which explain the passage of a speech program from the associated instantaneous value compressor 17 via the program selector matrix 23 and the gate circle oscillator 25 to the adder 26. The in the F i g. 3a at A shown waveform shows the processed audio frequency signal, which at the output of the instantaneous value compressor

17 appears. B shows the audio frequency sampling performed in each of the audio frequency sampling and storage circuits

18 is made. A uniform sampling of the instantaneous amplitude of the audio frequency wave A is carried out here at the beginning of one of the 128 ^ sec sampling periods. For example, time t _a denotes the beginning of a sampling period and t _b denotes the end of this sampling period and the beginning of the subsequent sampling period.

The audio frequency sampling and storage circuits 18 each have two inputs connected to the time pulse generator 24, so that the circuits 18 can be controlled for "sampling" and "terminating". For example, as shown in FIG. 1 a shows that the sampling pulse is emitted when the 8 pisec long 7.8125 kHz pulse F ₁₆ from the generator 24, which corresponds to the field 16, with the 1.33 μβεΰ long 125- kHz pulse, which corresponds to the time-division pulse I ₁ , coincides in time. The final pulse can be transmitted in a similar way if the pulse F ₁₅ corresponding in field 15 at the "And" gate 39 coincides with the time-division pulse t _e . The last-mentioned time partial pulse t _e understandably has the same duration and the same frequency characteristics as the pulses F ₁₆ and t _v
The in Fig. 3 a peaks rising above the line B are equal to the audio frequency wave amplitudes, so that these peaks embody the amplitudes at the sampling time, for example t _a and t _b.

13 14

At C , the outputs of the memory of the pitch modulation scheme of a pulse on an arbitrary frequency sampling and storage circuit 18 are shown. occur at the point of the sampling period. In the illustrated memory, the task of the example illustrated at B is to store the stored audio frequency, sampled amplitude up to the next signal positive, so that the pulse in question is stored in the sampling. As appears from waveform C 5 second half of the sample period. can be seen, the result is a staircase-like wave. It is now necessary to set the pulse positions in

to convert one of 16 specific time intervals to a sampled and stored variation corresponding to audio frequency wave A. which lie between t _a and t _b. The line / shows that

D shows a linear result of this conversion process as a function of time, which, as previously mentioned, increasing voltage, which is mentioned at each sampling period io, is carried out with the converter 20, after a reset with its initial value. In the example shown, the pulse that begins falls . This wave represents a steadily repeated line H in the converter stage 12 (which actually represents the sawtooth voltage. The period of this is the level between 11 and 12), so that the Um-sawtooth voltage corresponds to the 128 ^ sec sampler with this Impulse an 8 μ & εο long impulse period. The time wave is generated with the 7.8 kHz pulses 15 within the field 12, from the generator 24 in the pulse position modules. The output of the converter is generated around a converter gate 19. These 7.8 kHz pulses occur with a time period or a field delay, as indicated by line J at the same time as the "scan" commands for the. The purpose of this delay is to provide audio frequency sampling and storage circuits 18. The pulse from the pulse position modulator in a ge-time wave begins with a negative voltage level, the ao desired field over the entire 8 μβεο is equal to the largest negative voltage that ken, even if the 1 ^ sec trigger pulse in the audio frequency input wave will appear near the end of this field. It acts can. The time wave ends with a positive voltage - here it is a short storage of the position of the voltage, which is equal to the largest positive voltage trigger pulse. Since in the example shown is worth, which fell with the help of the audio frequency input 25, the trigger pulse in the field 12, can be achieved with the help of wave. In the illustrated representation of the converter 20 in the field 12 of the line J, despite the fact that an 8 μβεο long pulse was generated for the audio frequency wave A, a constant bias delay. Every word added. This bias voltage appears in the sampling period of each voice channel therefore contains the representation D as the zero voltage axis. The association 8 ^ sec pulse, the position of which is a converted use of a common timer input, ensures that the amplitude of the audio frequency signal is measured to ensure that the pulse position modulators are at the beginning of the special sampling period. Generate the period between the start of the scan and the end of the scan only once in each of the 128 ^ sec periods for a pulse. emerging 8 ^ sec pulses from the converters are in F i g. 3b at E , the combination of the stored program selector matrix 23 is supplied in the previously described audio frequency sampling with the time wave D 35.

shown. However, the time scale is significantly larger. In addition to the inputs from the 19 language ranges; it therefore corresponds to the area between the programs and the data program receives the lines AA and BB of the 128 ^ sec sampling period, program selector matrix also a series of, for example, the period between t _a and t _b in the outgoing from the time pulse generator 24. Representation A. The Lines F and G mark the occurrence of the time pulses. The pulses F are the 24, which appears at each of the six inputs of the series of termination signals for the Tonfrequentenabtast- and after from ^ to t _e and every 8 μβεΰ hold circuits 18 (Fig. 1 a). The pulse G is repeated, is shown in an enlarged view at K of the drawing the beginning of the return of the saw tooth Fig. 3c. These are the pulses t _v t ₂ , t ₃ , t ₄ , t ₅ and t _e in the time circuit of each pulse phase modulator 19 45 , which between and also the control command for the audio frequency time stamps t _c and t _d - 8 μβεο occur. The Koinzi sampling and storage circuitry for storing new denz in the program selector matrix 23 between the samples. 8 ^ sec pulse of every language program and

The enlarged representation of the wave train E in this group of six partial pulses is shown in FIG. 3 b shows a sawtooth of the sawtooth curve used in 50, in each language program the sequence of 1.33 sec-pulses previously fixed with a corresponding memory value on the five of an audio frequency sampling C. If the amplitude output lines of the program selector matrix 23 to the sawtooth generate the stored audio frequency sampling, which reaches the gate circuit oscillators 25 supplied value, a trigger circuit generates a short be. Each of the gate circuit oscillators generates a pulse of about 1 ysec duration. This pulse is in 55 another frequency, which is shown with f _v f ₂ , f _s , / ₄ and / ₅ in row H. It can be seen that they are drawn.

Location of this impulse is a function of the sound from the representation L in FIG. 3 c can be seen,

frequency wave stored amplitude. This that the particular time sequence within a link pulse would voted 8 μβεο long field between t _c and t _ä, in another site are located at a different amplitude at. If the moment at which the 1.33 \ isec long gate impulses for the five amplitude were equal to zero, this narrow oscillator would appear, through which the speech channel is determined at impulse in the middle of the sampling period t _a to t _b , from which find the original 8 ^ sec field. When the stored audio frequency sample pulse arrived. This time sequence arises in which it would be negative if this pulse were in the first program selector matrix 23. For example, half of the sampling period would be generated. If the sampling program 1, as shown at L , were positive in the time column I ₁ , the above-mentioned pulse would be a pulse at the input of the gate circle oscillator f _{v of} the second half of the sampling period. Thus in the time column t ₂ no pulse, in the time column t ₃ with this continuously operating pulse, a pulse at the input of the oscillator / ₃ , in the

15 16

Time column r _{4 shows} a pulse at the input of the oscillators Z ₁ and / ₅ , which are also typically tor f ₅ , in time column t ₅ a pulse at the input for the other three oscillators /. “/ ₃ and / ₄ and des Oscillator / „and in the time column t _e a pulse of the associated logic circuits at the input of the oscillator / ₄ . As mentioned before and in Figs. Ib

pointed out that this sequence of time and frequency 5 and 3 represented, five carrier frequencies Z ₁ to f ₅ are used according to the present invention of the coding or the signature of the program 1, corresponding to. For further explanation, each of the 19 converters is also arranged at L in such a way that it shows the pulse information coding corresponding to the particular message selected for the speech program 2. Here it can be seen that Z ₁ feeds in tion to each frequency channel. Each language proder time column t _v / ₄ in the time column t ₃ , / ₂ in the io grammkanal contains five "or" gates, which receive the time column t, _p / ₅ in the time column t ₅ and f _s in time converter inputs. Each translator has to appear with column ί _β . According to the present invention, these "or" gates of each frequency channel, every other language program has one sub-connected.

different sequence of time and frequency, as shown in the drawing sheet does not give the possibility of

Tables 1 and 2 shown above. 15 all frequency channel components of FIG. 4 plus

The output voltages of the oscillators, which are, however, always occur in this drawing in detail when release _{pulses appear at the logic circuits of the frequency channels f x} and / _s gate circuit oscillator inputs are shown at M. As can be seen, all converters are shown. For this reason, the emission follows certain "or" gates of each carrier freder high-frequency energy connected to this individual oscilloscope. The converter 2O ₁ has the same time-frequency scheme as the "Or" gate 41 of channel Z ₁ and the "Or" gate, the enable 54 of channel / ₃ appearing at its inputs and also with the other three not pulses. These five outputs are connected in an "or" port, shown in line, which combines the Frearen adder 26 shown in FIG. 1 b shows. quenzen / ₂ to / ₄ serve.

The combined output of this adder is at N 25. Similarly, the converter 2O ₂ with the types shown in FIG. 3c shown. It should be noted that these "or" gates 41 and 55, the converter 2O ₁₈ with the output is a carrier frequency that differs from an "or" gates 45 and 54 and the converter 2O ₁₉ with the 1.33 nsec long time intervals in which the "or" gates 45 and 52 are connected at the end. This handling of the gate circle oscillators 25 a signal impulse setters are understandably also with the was present, to the next of the 1.33 μεεϋ long 30-eared "or" gates of the channels, not shown, changes time intervals abruptly. It should also be connected to / ₂ , / _3, and / ₄ .

it should be mentioned that the outputs shown at JV "Da for each channel to receive the 19 Sprachder Language programs 1 and 2, despite their superficial programs, only have five "or" gates, lent similarities are actually unequal, since they are three or four with each "or" gate Time frequency order of these broadcasts associated with different language programs. This can be seen from FIG. 4 and recognize this difference on the receiver as in which programs 1 to 3 with the Identification of the two different language programs "Or" gate 41, programs 4 to 7 with the gate gram is recognized, as will be shown below 42, programs 8 to 11 with gate 43, etc. should be described in more detail. are connected. The remaining programs are

If, for two programs, the converted PPM-40 similar groupings and connections with regard to pulses appear in the same time field as the associated "or" gates, for example in field 12 of line /, two inputs L 20 of the data program can be required

Gate oscillators are switched on at the same time. no "or" gate, since the coding is as before. For example, programs 1 and 2 are described in connection with table 1. From the representations L and M that in the third time column 45 the data inputs L 20 lead immediately the two oscillators / ₃ and / ₄ work simultaneously. bar to the “and” gates that each time column ensure the corresponding receiver - the five frequencies are assigned. For example, for the necessary separation, which is necessary for the frequency J _1, the data input leads directly to differentiate between the two language programs 1 bar to the "and" gate 62 and for the frequency / ₅ un- and 2. 50 indirectly to the "and" gate 76.

The use of six time columns and five The language program information is used by each

Frequencies instead of only five time columns for an “or” gate to an associated “and” gate. This number of frequencies leads to a noticeable. The frequency Z ₁ are six "and" gates 61 loaned extension of the program coding. By being assigned to 66. Five of them receive by leaving a time slot free in each program, and if inputs are generated from the "or" gates, as mentioned above, there is a considerable increase in the number while the "and" gate62, as mentioned above, that of the possible Programs. A program can also receive data programs from this frequency channel. _{The frequency channel / 5 is} generated in a manner similar to that without the possibility of cross-talk. Such a program is used for the data transfer "and" gates 71 to 76, where the "and" - transfer. 60 Gate 76 receives the data program and the remaining

In FIG. 4, an abbreviated but detailed “and” gates the outputs of the “or” gates are shown in the scheme of the program selector matrix 23. Because the five "and" goals for everyone. In this representation, the converters 2O ₁ channel are effectively receiving information from three or four and 2O _{2 of} the language programs 1 and 2 and also the language programs, the possible converters 2O ₁₈ and 2O _{19 of} the language programs 18 and 65 speed of undesired crosstalk while are shown at, which embody the typical converters assigned to the 19 language, the sixth or data “and” gate of each program. Frequency channel this crosstalk does not occur. Furthermore, details of the logic circuits for the can occur.

909 512/1344

17 18

The "and" gates 61 to 66 of channel f _v restrict the transmitted frequencies in order to reduce "and" gates 71 to 76 of channel / ₅ as well as the "and" noise and interference to a minimum gates of frequencies f. -, up to / ₄ are arranged so that push down. The bandpass feeds the signal to it from the time pulse generator 24 Zeitspalteninforma- the mixer 112, which receive the signal with the functions, energy to the predetermined Zeitfre- 5 local oscillator vorzu- 113 coming Hochfrequenzquenzcodierung for each of the 20 programs as in the Heterodyne reception common, hear it in connection with F i g. 3 c and which is mixed. For example, the received panels 1 and 2 were discussed. If an 8 ^ sec signal in the frequency range between 4.4 and 5 GHz is an impulse from an "or" gate, for example from. The mixer 112 and the local oscillator "or" gate 41 at the associated "and" gate, in io 113 put this frequency in a range between this case the "and" gate 61 , a 55 and 65 MHz is converted .

1.33 ^ sec-impulse to the output "OR" gate 77 A preamplifier 114 compensates that in the mixer; This happens in the time gap, in the losses occurring in circle 112 , and makes the signal at the "and" gate 61 also a 1.33 ^ sec time pulse large enough for the multiple coupling circuit 115 to be connected. In the aforementioned case, it is the 15 dishes. The multiple coupling circuit feeds five time gaps t _r Die!, SS ^ sec output pulses to the intermediate frequency filter amplifier 116. Each of these "and" gates are passed through an "or" gate 77. The IF filter is at the frequency of one of the five and then fed to a pulse amplifier 79, different partial pulse frequencies matched, with which a so-called release pulse of the bandwidth is sufficiently wide to generate the corresponding 1.33 \ kSQc duration, which is used to switch on the ao chenden partial pulses, associated gate circle oscillator 25 . In a similar way, the intermediate frequency used can e.g. Am

Manner produce the other logic circuits / ₂ to / ₅ the above-mentioned 55 to 65 MHz range, such oscillator enable pulses when the IF filter logical in that at the output 116 appearing consequences see circles provided coincidences occurring five frequencies that lie in this range. The "and" gates 35 conduct in the frequency channel f ₅ and have the same frequency spacing as their outputs to the "or" gate 78, the output of which was present in the transmitter. outputs in the pulse amplifier 80 in release pulses. The amplitudes of these pulses fluctuate

being transformed. The reason for the tropospheric scattering according to the

How this is done in detail is shown in FIG. 5 law of chance. These fluctuations are more understandable because the pulse information is mainly of two different types, which appear in the 16 fields of a typical, namely a rapid change with a frequency of 128 ^ sec sampling period. According to the will- of about 10 Hz and a slow change, the pulse program of FIG. 2, is with periods longer than 10 minutes. In the Empdie F i g. 5 designed so that the code of each catcher is the fast randomly occurring special speech channel with the position of the pulses in 35 signal fluctuations that up to a certain F i g. 2 matches. For example, there appears to be a difference in the degree from one partial pulse frequency to the other in the field 1 of the period selected as an example in FIG. 2, effectively compensated by a pulse in program 10. Thus, this also shows squaring, integrating and summing. The EinFeld 1 of Fig. 5 l, 33 ^ sec impulses which are detected in the flows of the slow randomly occurring fluctuation-characterizing scheme of the program 10 with the help of an automatic handling. Strengthening controller 136 compensates, as in the As a consideration of Fig. 5 shows the Fig. 6b can. The input for these auto fields 6 and 11, in accordance with F i g. 2, automatic loss compensation 136 is the average value of the output of the two programs, formed for a long time over a period of some fields, and pulses from three adders 120 ₂₀ to be considered later, while some fields are not included. The one or more programs included. For example, the output of the automatic gain control holds the field 10 selected for explanation, which 136 is transmitted via a line A in FIG. 6 a all pulses of programs 4, 7 and 17. IF filters 116 are supplied. Here, a Kompen-It should be noted that the coding sation in that the gain of the filter of each program is prepared such that 50 amplifier 116 inversely detected for the combined is not varied more than a case in any of the signal.

128 μβεϋ lasting sampling periods of any two of those formed in the present system

Programs occur in the same time column. 20 channels are used, 19 for speech transmission. Pulses from programs 7 and 17 occur in the first, while one channel of data transmission uses time gaps in field 10 , but this results within 55 serves. The data transmission channel not only needs half of the 128 ^ sec sampling period in no other in any of the 16 time intervals to receive its impulses - time column further coincidences between the impulses, as previously described for the language programs of the two programs mentioned. because he can turn his impulses into one

It should now be based on FIG. 6 a and 6 b record any time interval. This data Detailed description of the receiver for the transmission channel 60 will be described below, given tropospheric scatter signal beam device The outputs of the IF filter 116 are using

will. The input to the receiver appears by five squaring circles 117 squared. The task of a parabolic antenna 110, which would give this squaring arrangement a bandpass, is to feed five, which generate the received high-frequency energy signals, the mean value of which is proportional to that which in this case is the shape of a multiple of energy, which in the corresponding has frequency pulse signal, which was available from the transmitter IF filter output. The squarer 117 of the system comes. The bandpass filter 111 has the open- are devices that would give instantaneous output signals that generate the bandwidth of the received signals, which is proportional to the squares of the

19 20

momentary input signals. The through-level at the output of the gate circuit integrator in the case of interference intersection value such a signal is proportional to the small noise.

Energy that is contained in the signals at the input of the Quadrie- When describing the transmitter was called Beirers. When using the quadric game, the language program PI considered more closely, it is not intended that it actually 5 and it was, as can be seen specifically from FIG. 3 c, performs an averaging. The squarer should use a certain scheme of frequencies and time rather only split the currently squared signals in order to generate one of its own for this program. The mean values themselves, on the other hand, will ultimately create coding. As the F i g. 3 c to the input circuit integrators formed because shows an image of this Fredie squarer ₁ value formation appears during transmission have no members for the Central io quenzschemas at the outputs of the IF filter Z. Overall, the combination of up to Z ₅ results in the corresponding time columns. Each of the squarers and the pre-circuit integrators, the filters, has a function of demodulators fed with squared signals. This squaring gate circuit integrator, as it was previously for the IF filter Z _1, work like a detector circuit in this regard. has been described. The outputs of this gate circle-The averaging, which is necessary to measure these energies, 15 integrators 118 appear in each time column am gien, is carried out with five gate circle integrators input of the program decoding matrix 119. These gates 118, each of which checks one of the qua-matrix let through integrator levels are assigned three. The integrators are switched on with regard to the correct time and frequency order, the beginning of each time column (partial pulse) in order to switch the language program FI when it occurs and to recognize again 20 at the end of each time column and switch off the output channel for the program, gramm PI for all other programs too

The aforementioned reception process should now lock out.

be described in more detail. In the troposphere the outputs of the gate circuit integrators are available via ren scattered radiation receivers, a certain program generates the lines B to F with the program selector matrix program, for example the program PI, which is connected, which is shown in FIG. 6 b. at the input of the IF filter / _} to Z ₅ a PPM pulse. This matrix has the same general structure, at the output of this filter in the special as the program selector matrix of the transmitter. The len time column uses a 1.33 ^ sec pulse with IF energy on the receiving-side matrix "and" - and as an identifier for this program. Thus, there are "or" gates that are similar to the previous output of a filter, for example filter f _v 30 in FIG. 4 described. The program selector from a 1.33 μβεο continuous transmission of the matrix 119 is the inverse counter-IF energy within the 8 μ $ & ο long field time. piece of the program selector matrix 23 of the transmitter. A group of five quadrilaterals of the squaring arrangement 17 is fed to the associated quad input of the matrix 119 as this signal emission. A new signal is generated in pulses that originate from the detection of the five frequencies in this squaring circle and are transmitted via lines B to F , which is and will always be positive due to the squaring. The output of this matrix is a group with an average DC voltage value equal to the average of twenty different signal programs. Six lower power of the IF input has. The corresponding 1.33- ^ isec time pulses I ₁ to t ₆ from circle follow the squaring-like 1.33- ^ isec time pulses from the associated gate circuit integrator. This the time circuits 133 cause by coincidence with the gate circuit integrator of the arrangement 118 at the beginning of the five input pulses from the squarers 117 start of the time gaps containing the IF signal and the gate circuit integrators 118 the distribution of the output levels. As a function of incoming programs, the integrator builds an output voltage on the corresponding time, that of the inte output channels.

grated area under the squared output signal To the subsequent operation of the speech proportional is so that effectively the total area 45 detectors and understand the data detector better Below is a detailed radio squarer below the averaged direct current pulse from the is measured. At the end of this 1.33 description of the aforementioned distribution pre-Usec time column the exit of the gate circuit appears given.

integrator as a voltage, which should be the area under It z. B. the time column t _± can be considered. to

is directly proportional to the squared signal 50 at this point in time, the matrix 119 receives via the

and thus a measure of the energy content, line G represents a 1.33 HSeC ImPuIs. This impulse has

because the integration of a power function according to the result that the energy detected _{at the filter Z 1, which}

Time leads to an energy measurement. the matrix 119 is reached via line B , with the aid of

This voltage, proportional to the signal energy, of the "And" - and "Or" - belonging to the line B is fed to the program selector matrix 119 of FIG. Grams 1, 2 and 3 are now to be transferred, as Table 1 shows, provided that no other programs are specified. It should also be mentioned that the IF filter output is present in arbitrary chart signals, in time gaps I ₁ at time I ₁ in channel Z ₁ , for example the output of filter Z ₁ may be present during the signal or not present in the other five aforementioned time gaps considered 60 so that the speech summers become one span. The output of the filter Z ₁ then consists of the received voltage which is proportional to either the signal energy or the random noise energy that occurs according to the law of chance. see. In the squarer, these interfering signals are The logical structure of the matrix 119 is selected, of course, squared, and it is also ensured that no other gate integrator is connected to the line B of the filter Z ₁ in each of the five time slots 65 program summer. With these measured the energy content of each interfering signal. However, the programs have similar Verbin-compared with the output in an actual fertilize f from the filter. ₂ to line C, from the filter ₃ Z signal is of course the DC to the line D, the filter Z ₄ to the line E or from the

21 22

Filter f _s to line F, as is required according to table 1 for each of each MWD 121 a box-car-shaped tone of these other programs. frequency wave that of the sampled, converted and

The other subsequent time gaps t ₂ to t _e, the audio frequency distributions fed to the transmitter of the system are the same depending on the requirements of the program wave. In other words, the table 1 encodes the five inputs to the rieh- 5 at the end of each period with corresponding Abtigen summers in exactly the same way as a voltage was written for the time column t _t reading and storage devices. Thus, at the end one stores is attesting to the arrival time of the correct im-sequence ?! to r _e , which is of course proportional to a time field pulses, so that with the outputs of 8 μ5 & o , each of the summers 12O ₁ to a sequence of sampling periods corresponds to a rising wave and 12O ₂₀ in five of the six time columns to a wave that descends from the magnitude is generated, the signal voltage impressed on a reproductrix, in order to generate an output of the original transmitter signal at a pulse whose voltage amplitude is proportional to the specific point of the original modulation front-end five voltages. ganges corresponds.

It can be seen that in the case of pure interfering signals the box car-shaped audio frequency wave is in

this summed voltage is low compared to the 15 audio sections processed by the MW the value that would result if detectors 121 were connected downstream in each. With this front five time gaps a real signal is present, especially the dynamic amplitude being were. If an overrange of the original audio frequency signal with speech pulse and in the other four time columns an expanding circle 122 restored. A Noise occurs, the result is similarly 20 band filter 123 eliminates the high frequency components Output voltage, the still considerably lower th of the box car-shaped shaft and the influences is than that in cases of correct signals. of the scanning process. That eliminates the pre-lifting

More specifically, the energy generated in each of the equalizer circuit 124 represents approximately the number five Partial impulses from the five gate circle integrato- quence-amplitude relation of the original Tonren 118 detected and summed in the summing circuits 25 frequency signal again. The audio frequency amplification, an output pulse, the amplitude of which the ker 125 amplifies the audio frequency signal level off-energy sum each of the five partial impulses proportional to a telephone line or a tontional is. In a time interval in which a special frequency transmitter 126, such as. B. a handset, Program impulse is expected, can - given a loudspeaker or a sound recording device, by the nature of the information transmitted - a 30 supply.

such impulse also fail. At such a time- the program selector matrix 119 feeds that too

point, the summing output pulse voltage is proportional to the data program 20 into a data program glitch and interference pulses. For a totalizer 12O ₂₀ a. The output of this summer speech transmission appears once within each is checked with a threshold detector 137, the sixteen pulse periods a correct pulse. 35 whose threshold value corresponds to the special off-thus result as the output of the summer 120 formation of the system set and a function of a sequence of correct pulses that occur during the still acceptable error frequency and the desired one sixteenth of the total period, and the th system threshold. The purpose of the threshold value additionally according to the law of the random occurring detector 137 is to determine the output level of the smaller glitch and interference pulses. 40 totalizer 12O ₂₀ at the end of every 8 ^ sec time

The components 121 are to check the field and make a decision about it Detectors detecting maximum probability of dropping whether a signal is present or not. at hereinafter referred to as MWD, and pulse position demo presence of a signal is indicated at the output of the impulse, subsequently also as a PPM demodulator pulse generator 138, a new or restored one ren called. The MW detectors of the component generate 45 8 ^ sec pulse. This output pulse forms tes 121 store the amplitude of all correct Im a marking signal. In the absence of a signal within pulse, all interference pulses and all interference pulses of an 8 ^ sec time field are generated by the pulse generator in each sampling period, then a decision 138 no output signal, corresponding to a Lükzu that do not meet a threshold value, but rather a ken- or gap state. Thus arises on is based on the premise that with a 50 pulse generator 138 a stream of characters and equal to all space pulses received during a sampling period, that of the original nen pulses correspond to the data pulse stream containing a message at the transmitter. Normal impulses characterized by a larger amplitude. wisely, these are time-subdivided multiplex units Step-like time wave serves as data in every MWD which, depending on the requirements, can be processed with a data-specific time scale to identify the processing device, arrival time of the correct pulse with respect to the For example, the data processing can begin the sampling period. This stair-like wave creates a demulti-solution that works with time division their task be by generating an output plexer, which is a group of teletype voltage, which is proportional to the time of arrival, d. i.e., connections. The function of a demultiplexer a maximum pulse which is late in a sampling 60 is the reverse of that of a multiplex device, period appears, produces a small output, i.e., the single signal that is multiplexed during a device arriving early in the sampling period is made up of multiple components Pulse generates a proportionally larger output, is again in Komposofern in the demultiplexer Within this period, not one impulse is resolved for individual channels and the larger one Amplitude follows. This voltage is fed from 65 speaking outputs, the end of the one sampling period until the end of the

following sampling period when the next totalizer 12O _{20 occurs} . The output sampling voltage is saved. Thus the output of this summer is a combined signal in

23 24

a multiple band of the fifth order. In this way, as can be seen from Table 1, we struggled to delay the rapidly changing fading phenomena, which correspond to the reciprocal values of the data program with the scattering in the troposphere switched off. These five impulses are reduced or substantially reduced. The slowly changing instantly added in the summer 129 to the fading phenomena, which also generate a 1.33 microsecond pulse with the scattering 5, the amplider troposphere of which, however, result in the sum of the detected individual pulse amplitudes the mean value of the combined signal with is proportional.

slower frequency changes. To get here a compen- Because of the result of the use of the five

To meet the sation, the previously discussed signal, which results in the summing frequencies, is used for an automatic gain control. Multiple transmission of the system is followed by the amplification or fading control in circuit 136. A control of this sum pulse is not the normal voltage signal, which corresponds to the mean value of the summing unit - Rapidly changing fading in the scattering in output signal is proportional to the auto- the troposphere; rather, the pulse retains a matic amplifier control 136 to the IF filter - relatively constant amplitude. The amplitude fluctuation amplifier 116 fed back in order to obtain the desired 15 kings by only slowly changing fading control. The degree of amplification of the phenomena is reduced by the previously described IF filter 116 is reduced with conventional amplifier-controlled fading compensation in the receiver, and changed. It should also be mentioned that this sum pulse,

It has already been explained that the gate circuit integrator hereinafter referred to as synchronization pulse gates 118 is released on the basis of time signals from the time 20, with a repetition frequency of 125 kHz 133 at the correct points in time, since it occurs in every δ-μβεΰ- ΖεϊίίεΜ appears and will be. It is evident that an exact syn- is used as the data pulse. If the data synchronization between the transmission-side time circuits is carried out arbitrarily, and the corresponding receiving-side time circuits do not need a pulse in every 8 ^ sec basic period in order to ensure proper actuation. However, the a priori probability of the occurrence of such a pulse in the various signal channels of the gate circle integration is 50%, gates and the program selector matrix allow- so that on average in every second 8- usec period. In order to ensure this exact match, an impulse appears. A spectral analysis such as this is intended to interlock between the reference signals interspersed at 125 kHz and time clock 135 of the receiver and the reference time clock 30 even with integer multiples thereof of the transmitter. This is used - as previously recognized tralines.

described - the frequency and phase information, no

those in which, according to the law of chance, crosstalk through impulses of the nineteen languages Data pulse train is included, which occur over the channels, since the data code is related to the Data channel is transmitted. 35 other previously described code is exclusive.

From FIG. 6 a is the mode of operation of this syn- In order from this synchronizing pulse stream the fre-

to recognize chronising circuit. It is first to take the sequence and phase information, the mode of operation of the system in normal operation is described phase detector 134 is provided, which is a ben here. It is assumed here that a syn important function has to be fulfilled. The phase de-synchronization has been achieved so that later on the 40 gate receives the initial call procedure at its two signal inputs, synchronization pulse stream and a similar one can be entered. For the synchronization method contains a special 125-kHz pulse current from the time circuits 133. The derter signal channel is provided to the Torkreis- phase detector generates always an output envelope detectors 127, the timers 128 and when the two input signals are not in phase, the summer 129 belong. The outputs of the five are 45. The polarity of the output is a function of IF filters 116 with respect to their envelope in the polarity of the phase difference. When the gate circle envelope detector 127 is checked in this way, the error signal is used to adjust the direction that is switched on or opened in accordance with the 750 kHz time circle 135 used for frequency and phase in accordance with the coding used for the data that the error signal the. A gate switch 132 ensures this switch-on 50 is reduced in size. When the required time signals are received from the extremely stable (quartz crystal controlled) time circuit, time circuits 133. At the output of each gate circuit with a variable-voltage fixed capacitor, envelope curve detectors 127 appear in each of the five in the oscillator circuit. These variable voltage output lines 1.33 μβεΰ DC pulses. The capacity is used to reduce the small amount of time required to achieve this. It is the frequency and the phase of the timing circuit desirable to take these five pulses into a temporal.

Bring coincidence. This is done with the help of the The 750 kHz output of the clock circuit 135 is a

Timers 128, which are responsible for delay times, which are used as a reference value for the timing circuits, which are used for the relative delays in the five frequencies, 133 which supply the basic time signals which are inversely proportional for frequency channels. For example, proper operation of the receiver _{gate will require the pulse from filter Z 1} by five pericircles.

dull, the impulse from the filter / ₂ by four periods, the It can now again go into the achievement of an impulse from the filter / ₃ by three periods, the impulse-catching synchronization. It from filter / ₄ by two periods and the pulse from 65 should _{delay the state when filter / 5} is switched on for the first time by a period of 1.33 usec. This delay scheme brings the five impulses exact synchronization between the transmitter and the Empin temporal coincidence, as this delay scheme is present. It is obvious that among these

909 512/1344

25 26

Conditions of the gate switch 132, which the gate during the normal operating period, ie after

has to turn on circular envelope detectors 127 , it is also desirable for the call to occasionally not delegate its work at the right time through a similar scheme in order to safely carry out. For this reason, the gate switch 132 will ensure that these reference values are maintained, except for 5 because of the overall synchronization of the tropospheric operation set during this synchronization call phase. In other words, that see scattered radiation systems are the gates at the transmitter and at the clock circles used at this point in time the Torkreishüllkurvendetk- receiver particularly stable gates are continuously open. The transmission side is and keep their frequencies for periods of approximately the transmission of speech or data as long as 10 seconds without a correction being necessary, until synchronization is established. io will. For this reason, occasional strong fading phenomena, which are referred to as the call period and which occur simultaneously in all five, a constant frequency will occur in the data program channel for periods of about 10 Se signals, which are periodically tolerated for subsequent without a new call being interrupted for reasons still to be explained.

will. For the greater part of the time, therefore, a 15 The above functional description of the excessive output signal is achieved on the basis of the transmission system shows that a single time data channel is supplied to the input. At the receiving end, the frequency coding scheme is used, in order to enable the superposition of a stream of 1,33 ^ sec pulses, which with a re-multitude of signal programs, results at the output of the data summer 129 to have the same basic time intervals. A recovery frequency of 125 kHz will occur. At this ao complete tribute and a fuller VerZeitpunkt the phase angle of 750 kHz Uhrkreises's understanding of the present invention result 135 incorrect leave for a frequency coincidence of the targeted benefits be obtained at a glance-time signal with this 125-kHz pulse current . However, the direction of the actual phase detector with the invention causes the work done with the 125 kHz device. One aspect of the synchronization pulse stream at one of its inputs, as is, for example, that in the case of a signal transmitted in a given language in the required synchronization program, the clock circuit is forced to display the. When the synchronization is achieved implemented in the program mentioned at the time, the gate switch 132 generates the timely correct amplitude level release signals for the gate circle envelope decoder - the latter according to the known pulse code modulators 127. This synchronization state is converted by the 30 tion technology into a pulse code wor synchronizing circuits checks which one is calling. However, while generating the usual pulse signal that turns the gate switch 132 on, code modulation technology uses the number of possible code combinations to switch the system from the calling period to the working code to approximately the number of overtime. Subsequently, voice and data can be transmitted in standardized information steps or in a painterly manner, 35 only the addition of a few extra, whereby the synchronization is retained in the previously described steps for equality checks possible. makes are effective in the system according to the invention

Furthermore, in the call period, the positive must be present in every 128 μβεσ long sampling period at the beginning of the sixth of the speech sampling periods of 128 μβεο duration ninety time columns. The number of may be captured. This is necessary for the proper encodings, with ninety-six steps the demodulation of the converted pulse position 2 ⁹⁶ , ie effectively very large. From this large modulated speech signals. The identification number of coding possibilities uses each program pulse, which of the program selector matrix is only sixteen. The error lying here is required, at the beginning of every language correction possibility are open to the day. To mark the sampling period, in the time circles 45, for example, a level can be generated as a number of NuI-133. In order to initiate this signal at the correct time, five consecutive ones and subsequent point with respect to the similar transmission-side reference region a number of zeros are shown so that time is, as previously mentioned, while the total number of zeros and ones is 96, the Call period at the sender in the continuous Due to the crosstalk and the permanent data input a scheme was introduced. This 50 avoidable noise can be a number of schemes, e.g. B. be a three-digit Barker code. Zeros appear as ones in the receiver. The Barker code is useful because it is known to deny the MW detector, but this additional crosstalk-free can be detected. This Barker code ones and can make its decision in the part of the scheme is made at the beginning of the sampling period on the total Sen code, which contains the five ones, which is inserted to subsequently in the synchronizing 55 That is in other words that the MW detection pulse stream at the output of the summer 129 of the gate an error detection and correction system to appear after the receiver during the call time. A scheme recognition circuit 131 of the receiver simplifies the arrival time of this Barker scheme in terms of its circuit structure and construction than a common error detection and correction circuit provided for the same purpose through continuous measurement of the autocorrelation of the incoming impulse current. The digital values. This special property gives the essence of the Barker scheme is recognizable by the possibility of over-autocorrelation of a Barker scheme in the same frequency band a sharp increase in amplitude, which characterizes the large number of signals over time.

is ending. In this way, in the time 6 ₅ While a preferred embodiment of the Erkreisen 133 a strobe pulse serving as reference has been described the invention, it is also with in the value for the start time of the sampling period scope of the invention, the basic objectives in various functions which other Embodiments to achieve. At-

for example, the number of frequencies used to improve diversity be less than described if less gain is needed. on the other hand the number of frequencies can also be greater than described if a higher gain is required will. A wide variety of Provide combinations of voice and data channels, and it can be used in the voice converter circles use a different number of conversion steps when higher fidelity or lower fidelity is required can be tolerated. Such changes only lead to changes in the bandwidth of the transmitted signal.

Different components and switching methods of the system can also be used in different ways be designed without going beyond the scope of the present invention. For example, ao Instead of the MW detector and the associated demodulator, a simple threshold value detector is used if only a low level of crosstalk pulses is expected in the voice channels. As a further example, it should be mentioned that other types of signal detectors are also possible, for example Detectors with matched filters or correlation detectors.

In the receiver, pre-amplifiers, for example parameter amplifiers, can be provided without going beyond the scope of the invention.

While in the described embodiment of the invention a pulse transmission system is applied because of the inherent advantages of this system in data processing and operation of multiple relays (multiple hop relays), it is well within the scope of the invention, the multiple frequency multiple system (multiple frequency diversity scheme) with a single transmitter-side power amplifier as well as an associated antenna and antenna feed and a single receiving antenna as well as an associated antenna feed and a receiver in addition to impulse systems can also be used for other signal structures. For example, in connection with a conventional frequency-modulated signal can be used with this technique. In this case, the Multiple oscillators work continuously and are no longer released by pulses. Everyone who Multiple oscillators would be of the same audio frequency signal according to frequency or also according to modulated in phase. Thus, the same frequency-modulated signal would be transmitted to all different, for the improvement of the diversity (diversity) provided frequencies are transmitted.

At the frequencies for which the tropospheric scattered radiation system is practically possible, lie the carrier frequencies are roughly on the order of 1 MHz apart, and it stands in each of the Multiple transmission carrier frequency bands of sufficient spectral space are available to the usual Perform frequency division multiplexing.

Claims (11)

Claims: 6s 1. Scattered radiation diversity transmission system with a plurality of transmission channels, which contains a single transmitting and receiving system and is equipped with a single transmitting and receiving antenna for each transmission device, characterized in that the transmitting and receiving side information converter (18 to 20, 23 to 28 or 116 to 120, 133) are provided, which encode the signals of all transmission channels according to a certain time and carrier frequency scheme in such a way that the message to be transmitted is modulated by means of basic signal pulses, which consist of a number of equal partial pulses (ti to t6) are divided, each of which is transmitted on a separate frequency. 2. Transmission system according to claim 1, characterized in that program selectors (23, 119) are provided for coding or decoding the signals of all transmission channels at the transmitter and at the receiver. 3. Transmission system according to claim 1, characterized by a transmitter for each Frequency contains a separate oscillator (25) and a combination circuit (26) in which the Signals before radiation at the antenna according to the time and carrier frequency scheme can be combined (Fig. 1 b). 4. Transmission system according to claim 1, characterized in that all transmission channels of the transmitter contain a pulse position modulator (19) and all channels of the receiver contain a pulse position demodulator (121). 5. Transmission system according to claim 1, characterized in that one of the transmission channels (program 20) of the transmitter contains a digital value encoder (34) and one of the transmission channels of the receiver contains a data processing device (139) . 6. Transmission system according to claim 1, characterized in that modulators (41 to 45, 51 to 55 in FIG. 4) are provided in the channels of the transmitter, which generate periodic pulses which indicate information of the associated channel, and that a generator (24) generates a plurality of partial pulses (^ ₁ to t _e ) for each modulator pulse, each of which has a specific temporal distribution for a specific information channel (Tables 1 and 2). 7. Transmission system according to claim 1, characterized in that each channel of the receiver contains frequency selection devices (119) which separate the received frequencies into groups of partial pulses, and that a further distribution device (120) is provided which each group of partial pulses in individual pulses for separates the individual information channels. 8. Transmission system according to claim 1, characterized in that a detector (121) operating with maximum probability is provided for each of the channels of the receiver and distinguishes between pulses which have been coded for this channel and crosstalk pulses which have a lower amplitude than the pulses encoded for that particular channel. 9. Transmission system according to claim 1, characterized in that the receiver has a clock circuit (135), a phase detector and a Synchronization circuit (134) contains the latter, the clock circuit with the clock circuit on the transmitter side (33) keeps in synchronism, the phase detector circuit and the synchronizing circuit (134) when synchronizing respond to the signals of the data channel (program 20). 10. Transmission system according to claim 1, characterized in that a shrinkage compensation (136) to compensate for according to the law the random rapid and slow fluctuations in the transmitted signals is provided. 11. Transmission system according to claim 1, in which speech information is present in some transmission channels are transmitted, characterized in that the speech information in each individual channel in step-like partial signals is implemented (20). In addition 4 sheets of drawings